Voice Identification, FRE 901(b)(5) And The Fifth Amendment Due Process Guarantee

The Seventh Circuit notes that voice identification is no different than any other means of identification for Fifth Amendment Due Process purposes; expert testimony was not needed to identify a drug conspiracy defendant’s voice on an intercepted recording introduced at trial; the case agent’s “minimal familiarity” with the defendant’s voice, based on listening to an intercepted conversation on the day the agent arrested the defendant and on speaking with the defendant the same day “falls squarely” within the FRE 901(b)(5) requirements for voice authentication, in United States v. Recendiz, 557 F.3d 511 (7th Cir. March 3, 2009) (Nos. 06-1754, 06-2380, 06-2821)

Voice identification is often the source of dispute. FRE 901(b)(5) explicitly allows a witness to express his or her opinion on the identify of a voice based on “hearing the voice at any time under circumstances connecting it with the alleged speaker.” This form of authentication is most often used to admit recorded conversations. It may also be used to tie a party to an incident or to allow a witness to identify participants in a conversation where the witness heard the conversation but did not see who was speaking. A recent Seventh Circuit case considered a challenge to an agent’s identification of a defendant’s voice.

United States v. Recendiz

In the Recendiz case, defendant Navar was charged as a participant in a large cocaine trafficking conspiracy. At trial, the government tried to show that the supplier of drugs known as “the Doctor” was the defendant. Specifically, “[DEA Special Agent] Tulshi testified that on the day of Navar’s arrest, he listened to Call No. 331, which was a six-minute conversation in Spanish between [cooperating co-defendant] Herrera and a man whose voice Tulshi did not yet recognize. Later that day, Tulshi, who speaks Spanish, participated in Navar’s arrest and post-arrest interviews and conversed with Navar. Tulshi testified that after hearing Navar speak, he recognized Navar’s voice as the second speaker on Call No. 331.” Recendiz, 557 F.3d at 526. Following his conviction, the defendant on appeal claimed that he was not “the Doctor” and that the voice identification “(1) lacked foundation, particularly because the trial court had no basis for qualifying Tulshi as an expert witness, and (2) was unduly suggestive in violation of Navar’s constitutional right to due process.” Recendiz, 557 F.3d at 526-27.

Application Of FRE 901(b)(5) Authentication Requirements

The Seventh Circuit rejected both of the defendant’s objections. On the first challenge that the identification lacked proper foundation, the agent demonstrated that he had the necessary “minimal familiarity” with the defendant’s voice. The agent’s identification of the defendant’s voice fell “squarely” within the FRE 901(b)(5) requirements for voice identification. The rule allows voice identification through use of lay-opinion testimony when the witness bases the opinion on “hearing the voice at any time under circumstances connecting it with the alleged speaker.” FRE 901(b)(5).

The rule’s requirements were satisfied by the agent’s testimony that “he listened to a recorded phone conversation between [co-conspirator] Herrera and another speaker on the same day that he arrested Navar. Later that day, he spoke with Navar during his arrest and post-arrest interview, each of which connected Navar’s voice to his identity. Tulshi then testified at trial that based on hearing Navar’s voice in person, he recognized it as the same voice he heard on Call No. 331.” Recendiz, 557 F.3d at 527.

The circuit rejected the defense contention that expert testimony was required for voice identification. The circuit recalled that the Advisory Committee notes to the rule specifically recognize that voice identification is “not a subject of expert testimony.” ACN FRE 901(b). Consequently, “expert testimony is not required to authenticate a voice identification.” Recendiz, 557 F.3d at 527 (citing United States v. Magana, 118 F.3d 1173, 1208 (7th Cir. 1997) (“‘As long as the basic requirement of familiarity with the voice is met, lay opinion testimony is an acceptable means for establishing a speaker’s identity.’”) (quoting United States v. Vega, 860 F.2d 779, 788 (7th Cir. 1988))).

Identification Procedure Was Not Unduly Suggestive

The Seventh Circuit also concluded that the agent’s voice identification was not “unduly suggestive in violation of Navar’s constitutional right to due process.” Recendiz, 557 F.3d at 526-27. The voice identification was not impermissible under the Fifth Amendment merely because it was made on “the same day” that the witness “knew the authorities were going to arrest Navar.” Recendiz, 557 F.3d at 528. Because the defendant presented “nothing beyond the sequence of events to indicate that the process suggested to Tulshi that Navar was a participant on Call No. 331,” the circuit found that the identification procedure used was not unduly suggestive.

According to the Seventh Circuit, under the Fifth Amendment Due Process Clause, even had the circumstances of the agent’s in-court voice identification been unduly suggestive, the agent’s level of certainty “connected defendant’s voice to his identity” so that it did not render the identification “constitutionally” unreliable. The circuit noted that even had the procedure been unduly suggestive, the agent’s identification would not violate the defendant’s right to due process under the Fifth Amendment. The identification procedure was reliable - the “linchpin in determining the admissibility of identification testimony.” Recendiz, 557 F.3d at 525 (quoting Manson v. Brathwaite, 432 U.S. 98, 114 (1977) (In assessing whether an identification was reliable, a court should consider the following factors: (1) the opportunity of the witness to view the criminal at the time of the crime (or prior to the identification), (2) the witness’s degree of attention during such an opportunity, (3) the accuracy of the witness’s prior description of the criminal, if he made one, (4) the level of certainty demonstrated at the time of the identification, and (5) the time between the crime and the identification.)). The circuit applied the “Biggers” factors based on Neil v. Biggers, 409 U.S. 188, 198 (1972), as the following chart demonstrates:

Biggers Factors Applied To Both In-Court And Voice Identifications

Bigger’s FactorIn-Court IdentificationVoice-Identification
First Factor: Witness's Opportunity to View DefendantThe witness had seen the defendant "over twenty times" before his arrest, which was an "ample opportunity" for identification.“Special Agent Tulshi heard Navar’s voice in person on the day of his arrest, and he also had a clear opportunity to listen to the voice on Call No. 331-the call lasted six minutes and was a recording.”
Second Factor: Degree of Attention to DefendantThe circuit found that the witness did not address his "degree of attention" during his meetings with the defendant.“Tulshi testified that he listened to Call No. 331 with the express purpose of remembering the then-unidentified voice. Knowing that as a special agent his recollection would be subject to close scrutiny at trial, Tulshi devoted proper attention to the call, making him nothing like “a casual or passing observer.”
Third Factor: Accuracy of Prior Identifications by WitnessWhile the witness did not provide a detailed description of the defendant prior to trial, he did make "a detailed description" of the defendant prior to the in-court identification, satisfying the third factor.“[B]ecause Tulshi’s familiarity with Navar’s voice came from a recording of the telephone call, the accuracy of the voice is clear.”
Fourth Factor: Level of CertaintyThe witness "exhibited no uncertainty at trial" as to who was the defendant noted the circuit, once he was allowed to view persons within the range of his visual focus.“Tulshi expressed no uncertainty that, after hearing Navar’s voice during his arrest and post-arrest interview, the voice on Call No. 331 belonged to Navar.”
Fifth Factor: Time LapseWhile the in-court identification occurred almost two years after the witness's last interaction with the defendant, "any concern over that length of time is diminished by the strength of the other factors-particularly Tmiri's familiarity with Navar and his specific description of him."“[T]he fifth factor also favors admissibility because Tulshi listened to Call No. 331 on the same day that he participated in Navar’s arrest.”

All five Biggers factors supported the admission of the agent’s identification. “Any remaining concerns regarding the accuracy of Tulshi’s recollection of the voice [wer]e relevant to the weight of the testimony, not its admissibility.” Recendiz, 557 F.3d at 528. In finding no error in admitting the agent’s lay voice identification opinion testimony in Recendiz, the Seventh Circuit emphasized the flexibility of FRE 901(b)(5) that other circuits have long recognized as well. See, e.g., United States v. Thomas, 586 F.2d 123, 132 (9th Cir. 1978) (Voice identification testimony was not “improper opinion testimony by a lay witness” because all that was required for admission of the voice identification was that the lay witness have a “requisite familiarity with the speaker.”).

Finally, the circuit concluded that the constitutional test for whether an identification is sufficiently reliable to withstand Fifth Amendment due process scrutiny (as set forth by Neil v. Biggers, 409 U.S. 188, 198 (1972)) is the same regardless of the form of identification used. Consequently, the same constitutional standard applied to voice identification.

The Recendiz case presented another evidence issue concerning an in-court identification, which was previously considered. In Recendiz, the circuit had applied these factors to approve an in-court identification by a witness having trouble seeing as harmless, even if it had been unduly suggestive. See In-Court Identification By Witness Having Trouble Seeing Was Not Unduly Suggestive.

Federal Rules of Evidence